1. Field of the Invention
This invention relates to digital optical recording and retrieval from data storage structures. More specifically it relates to three-dimensional mark, land and tracking guide configurations for improved Figure of Merit, and to methods and apparatus for production thereof.
2. Description of Prior Art
Information storage and retrieval employing a spinning disc upon which digital data (representing documents, software, music, images and other types of information) are recorded, and from which data are retrieved by means of one or more optical beams impinging on the surface of the disc (or a xe2x80x9cdaughterxe2x80x9d disc produced from a disc master), is well known in the art. Likewise well known are other structures upon which digital data are optically recorded and from which the data are retrieved, including cylinders and cards. Optical recording on multi-layered data xe2x80x9cwafers,xe2x80x9d from which data are selectively extracted by scanning or other means essentially free of moving parts, can easily be anticipated in the relatively near future, as well as other sophisticated structures for ultra large-scale optical data storage and selective retrieval.
While each such type of structure might be separately addressed in this discussion, it is believed that the concepts discussed can be more clearly addressed with particular emphasis on essentially planar, disc-shaped structures, upon which data are recorded and from which data are retrieved while the discs are spinning about a central axis. From time to time, reference may be made to these other structures in connection with the invention disclosed herein, whose application and embodiments are by no means limited to discs.
A number of commonly employed disc-based optical data recording methods exist, each proceeding on a fundamentally different physical basis, and various implementation variations exist within each method. However, these optical disc recording methods have a number of features in common. For example, they all utilize a spinning, disc-shaped storage structure upon whose surface (or surfaces, or layers) one or more spiral data tracks are imposed. In some applications, there may be only one continuous track on the particular surface; in others there are a plurality of tracks, each occupying an annulus on that disc surface. Of course, in essentially rectangular optical recording structures, such as cards and data wafers, the tracks would likely constitute approximately parallel lines of data marks.
Each data track comprises a succession of a great number of microscopic marks interspersed by unmarked, or differently marked, areas commonly designated as xe2x80x9clands.xe2x80x9d The track pitch (i.e., the radial distance between the longitudinal axes of adjacent, essentially circular track portions) is microscopic, as is the length and width of each of the marks. Accordingly, a data track on a disc surface may be thought of as a large number of closely spaced essentially circular pathways, each containing a great many data marks and intervening lands in succession. In some applications, a particular data track or track portion might not be completely circular, in the sense that it might occupy only an arc of a circle on the disc. However, in this discussion arcuate and circular data tracks or portions of tracks will be referred to interchangeably as being circular data tracks. Since the circumference of each of these essentially circular pathways is very great, in comparison to the dimensions of the marks and lands, a small succession of marks and intervening lands will appear to be a linear (i.e., straight line) sequence at the microscopic level. Accordingly, at the microscopic level, radially adjacent data tracks on the disc may be viewed as essentially parallel lines of data, each containing a longitudinal succession of linear marks and lands, although at the macroscopic level they are essentially concentric circular paths.
The disc is normally written (i.e., recorded) and read by rotating it rapidly on a motor-driven spindle. Trackingxe2x80x94maintenance of the radial position of the write beam and/or the read beam precisely in the center of the data trackxe2x80x94is accomplished through a servo apparatus that compares at least a single pair of continuous readings. Each reading in the pair is taken on opposite sides of the longitudinal axis of the track. Based on these readings, the servo continuously adjusts the radial position of the beam to cause the readings on opposite sides of the track to be equal. This condition occurs when the beam is focused precisely upon the longitudinal axis of the track, i.e., when the two reading points are equidistant from the track axis. The sensing method will, of course, depend on the particular optical data recording method employed.
Tracking may be accomplished with a single beamxe2x80x94the read beam or the write beam, depending on which operation is being tracked. Here, the reflected beam is optically split into a data retrieval beam (or write monitoring beam, in direct-read-after-writexe2x80x94DRAWxe2x80x94applications) and a tracking beam. In CD and DVD-R applications, the reflected tracking beam component of essentially circular cross-section is divided into two equal semi-circles, the dividing line between them being parallel to the longitudinal track axis. The tracking sensor continuously compares the intensity of the two halves of that image, and a servomechanism adjusts the radial position of the beam to cause the sensed light in both halves to be equal. The latter condition indicates that the readings are being taken from the center of the track axis, i.e., that proper tracking is occurring.
Generally, the same sensor is employed for tracking and for data retrieval (or write monitoring). In single-beam (xe2x80x9cpush-pullxe2x80x9d or xe2x80x9cPPxe2x80x9d) CD tracking, one of the two sensed components is subtracted from the other, and a zero difference (i.e., equal input from both sides) indicates proper tracking. Data retrieval (or write monitoring) is accomplished by adding the two halves.
In DVD-ROM applications, differential phase tracking is employed, in which the reflected light is divided into four quadrants and the phasing of each is compared to determine tracking condition.
Most CD playback devices employ triple-beam data retrieval tracking (although the xe2x80x9cRed Bookxe2x80x9d only prescribes standards for single-beam, PP tracking). In triple-beam tracking, the read beam is split into three beams, the read beam itself, a first tracking beam directed one or more mark lengths ahead of it and offset xc2xc of the track pitch (approximately xc2xe mark width) to one side, and the second tracking beam directed one or more mark lengths behind the read beam and offset xc2xc of the track pitch to the other side. Each of the two tracking beam reflections is individually sensed continuously for tracking, in the manner described above in respect to PP tracking.
Beam focusing is likewise accomplished through a suitable feedback mechanism. Beam focusing is commonly employed and well known in the art, and therefore need not be further described except as may be necessary to describe particular applications.
The specifics of optical data recording, retrieval and tracking depend on which type of optical data recording is being considered. Accordingly, to understand optical data recording and retrieval, and to further understand trackingxe2x80x94and the present inventionxe2x80x94it is important first to consider the various optical data recording methods commonly in use, with particular attention to those aspects of data retrieval and tracking pertaining to the present invention and its various embodiments and applications.
The most commonly employed optical data recording methods fall into four categories.
In Magneto-Optic (xe2x80x9cMOxe2x80x9d) data recording, the general purpose is to store erasable data files on a disc for archival purposes, e.g., in computer hard drives. In MO recording, the disc""s recording surface is comprised of one or more thin metallic-alloy layers having specific magneto-optic properties, which are sandwiched between thin dielectric and non-magnetic metallic layers. Together these layers comprise the MO data disc""s recording surface. This recording surface is normally applied onto a polymeric substrate, as discussed below.
The recorded marks that correspond to a binary pulse-length modulated waveform (i.e., the physical representation of the digital data to be recorded) are essentially two-dimensional. This is because these marks are produced in only one of the MO disc""s magnetic film layers, which usually has a thickness of only xcx9c0.05 microns (one micron being one thousandth of a millimeter), or less. The unrecorded MO disc is pre-grooved, i.e., the polymeric substrate upon which the disc""s recording surface is laid possesses a continuous, three-dimensional spiral groove having a width of xcx9c0.5 micron, upon which the track(s) of data marks in the recording surface will be formed. This pre-groove may be created in the unrecorded MO disc substrate by molding it from a stamper made from a disc master. The disc master, in turn, may be produced by a number of methods, including the PR and dye-polymer methods, which will be discussed below. The cross-sectional shape of the grove is a matter of choice, most being essentially trapezoidal. The track pitch is normally xcx9c1.2 microns. During recording and subsequent playback of a recorded MO disc, the tightly-focused recording and reading beam spots xe2x80x9cfollowxe2x80x9d this pre-groove by PP tracking.
A spatial replica of the temporal binary pulse-length modulated waveform that physically represents digital data is recorded on an MO disc in the following way: The material comprising the thin magnetic storage layer in an MO disc is selected to have two stable orientations: xe2x80x9cupxe2x80x9d or xe2x80x9cdownxe2x80x9d relative to the surface of the storage layer. Since the magnetic medium xe2x80x9ccoatsxe2x80x9d the entire disc surface, including all surfaces within the grooved track, the actual direction of the magnetic orientation depends on where it is measured, although it is always perpendicular to each surface. An up-magnetized microscopic region (mark) normally corresponds to a high voltage pulse of the binary data signal being recorded, where the down-magnetized region then corresponds to a low voltage portion. Of course, the up and down magnetic orientations can be reversed in data significance, where down-magnetized marks represent high voltage pulses and up-magnetized marks represent low voltage portions.
An MO disc is initially prepared in a single magnetization state. In other words, the entire disc surface is initially in one of the two stable, mutually opposed magnetization states. Recording therefore consists of selectively xe2x80x9cswitchingxe2x80x9d the magnetization of spaced-apart, constant-width, elongated microscopic regions of the thin magnetic recording layer lying within the pre-grooved track in the disc""s surface. These switched regions within the pre-grooved spiral track are the recorded marks. The interspersed regions within the groove that are not switched (relative to the initial state)xe2x80x94i.e., which remain in their initial magnetization statexe2x80x94are the lands.
Once recorded, the marks and intervening lands are read by means of a narrowly focused beam of light, normally from a laser source. Linearly polarized playback light that is reflected from a microscopic up-magnetized region of the recorded surface will be rotated in a clockwise direction, and the light reflected from a down-magnetized region will be rotated in a counter-clockwise direction. By detecting the polarization rotation in the light reflected from the disc, the read mechanism determines the magnetization orientation of the mark currently being read by the focused reading light spot, and thus retrieves the data value (xe2x80x9c0xe2x80x9d or xe2x80x9c1xe2x80x9d) of that mark.
Unlike MO data retrieval, which depends on detection of polarization rotation, MO tracking employs purely optical PP mechanisms, based on comparison of the intensity of reflected light detected in two halves of the reflected image from the tracking groove. The read beam is moved radially (relative to the disc) until the two halves detect the same reflected intensity, as discussed above.
Each mark written in the pre-grooved track of the storage layer of an MO disc comprises a large number of extremely minute magnetic domains that have the same magnetic orientation, which coalesce into the relatively larger microscopic domain comprising the recorded mark. The magnetic orientation of these minute magnetic domains within a given mark is set by an external magnetic field that exceeds the thin magnetic storage film""s coercivity. Accordingly, the magnetization of a domain in the thin magnetic film will align with the force lines of a sufficiently strong external field, if that external field is in one of the two stable magnetization directions of the film. The magnetic-alloy materials that comprise the data storage layers of MO discs have magnetic coercivities that are extremely temperature dependent in a non-linear manner. In fact, at room temperaturexe2x80x94indeed, at any temperature below the Curie temperature of the magnetic alloy used to form the storage layer (the Curie temperature is generally in the range of 200 to 300 degrees centigrade for the magnetic alloys commonly employed in MO discs)xe2x80x94unattainably large external magnetic fields would be required to switch the magnetic orientation of the storage layer""s domains. However, if the magnetic storage layer""s temperature is raised to its Curie temperature (or slightly above it), a very small external field can re-orient the magnetization direction of the tiny magnetic domains.
Therefore, to record MO marks, i.e., to orient the minute domains of the thin magnetic storage layer in an elongated region of the pre-groove on the data surface of the MO disc, a tightly focused spot of light is employed to locally heat the thin magnetic-alloy storage layer to just above its Curie temperature, while simultaneously applying a relatively small external magnetic field that is oriented either up, or down, relative to the plane of the disc. Causing current selectively to flow briefly in either of two directions through a nearby coil of wire that is wound parallel to the plane of the disc can create this controllable external magnetic field. Thus, the focused recording light spot merely serves to define the region of the pre-groove that may be affected (i.e., have its magnetization re-oriented) by the relatively small external magnetic field. Accordingly, it is the external magnetic fieldxe2x80x94not the focused light beamxe2x80x94that actually forms the mark in the storage layer of the MO disc.
Once again, the material comprising the storage layer is selected so that the imposed magnetic orientation on all surfaces within the pre-grooved track is xe2x80x9cup,xe2x80x9d relative to each surface, if the orientation of the external magnetic field is xe2x80x9cup,xe2x80x9d and vice versa. In all cases, the magnetic orientation is perpendicular to each such surface.
Because it depends on use of a light beam to locally elevate the temperature of the recording medium, in order to cause a small domain in the medium surface to exceed its Curie temperature, the MO method of optical data recording may be classified as a thermal optical data recording method. In other words, it is not the quantity of light, per se, that causes the desired effect; it is the heat locally generated within the surface of the medium that facilitates magnetic-orientation switching by quickly raising a small area of constant width above the Curie temperature of the material, so that the relatively small external magnetic field can effect the desired switching. Since the disc is rapidly revolving during the process, the created mark becomes elongated so long as the light beam remains on (and the magnetic field orientation is maintained). Thus, switching of the orientation of the external magnetic field, to represent the temporal binary data, in coordination with the application of the heat-producing light beam to selectively raise a group of domains in the surface to above the Curie temperature of the material, creates marks of selective length and magnetic orientation, interspersed with marks of selective length and opposite magnetic orientation. The width of the marks is determined by the beam width (see, below), and the width of the pre-groove is selected accordingly, in a manner familiar to those with ordinary skill in the art.
An MO data disc recording can be written and re-written as desired, and this feature is, of course, utilized in computer hard drives. Once the MO disc is written, the focused recording spot can be made to scan any desired segment(s) of the data track(s) while the power level of the light in the spot is maintained at a level sufficiently high to raise the recording medium to above its Curie temperature, while the external magnetic field strength and direction are maintained constant. This will erase any desired segment(s) of disc track(s) by uniformly re-orienting all the domains within such track segment(s) in a single direction (say, xe2x80x9cupxe2x80x9d). Once erased, the segment(s) may be re-written in the manner described above. This erase/re-write procedure can bexe2x80x94and normally isxe2x80x94applied in a continuous manner, to completely xe2x80x9cwrite overxe2x80x9d any desired portion of the MO data disc. There are many sophisticated methods of accomplishing this, which are commonly applied and well known in the art (particularly in reference to data storage on computer hard drives), but which are not directly relevant to the present discussion.
Since the magnetic storage layer employed in MO discs is quite thin, the in-plane thermal heat diffusion in this layer will be large relative to diffusion in the perpendicular direction. Furthermore, because this layer has some finite heat capacity, it takes time for the film to cool after it is heated to its Curie temperature. Therefore, a magnetically switched region in the thin MO storage layer will continue to lengthen (since the disc is spinning) after it has passed beneath the focused recording spot, since the layer remains exposed to the coercive force of the less tightly focused external magnetic field. To minimize such residual domain growth (which is commonly referred to as domain xe2x80x9cbloomxe2x80x9d), an MO disc may incorporate a metallic heat sink layer (which doubles as a light reflector) that is located close to the magnetic storage layer. This heat sink layer will pull heat out of the storage layer and thereby increase its cooling rate. Also, since the transverse extent of the formed mark is limited by the pre-groove""s sidewalls, bloom is mostly a problem in the longitudinal (along the track) direction. Therefore, it chiefly affects the correct length and xe2x80x9cduty cyclexe2x80x9d (ratio of mark to mark-plus-land length) of the written marks.
Since the magnetic storage layer in an MO disc is quite thin, heating to the Curie temperature begins to occur almost immediately upon switching the laser beam to its xe2x80x9conxe2x80x9d position. And because of the heat sink, the temperature of the medium almost immediately cools to below its Curie temperature when the beam is switched xe2x80x9coff.xe2x80x9d However, in practice the leading end of an MO mark is normally narrower, in plan view, than its trailing end. This is becausexe2x80x94noting the fact that the disc is rapidly rotating while mark writing occursxe2x80x94a small amount of time is required to bring the leading end to the Curie temperature, while the heat sink insures that virtually no time at all is required to cool the trailing end to below the Curie temperature. Since accurate data retrieval ordinarily depends on detection of the mark-land transitions, this means that the leading end of a mark should have nearly the same shape as its trailing end, so that the mix of xe2x80x9cupxe2x80x9dxe2x80x94xe2x80x9cdownxe2x80x9d orientation in each transitional area (i.e., the leading and trailing regions of the marks) may be defined and consistent.
Accordingly, the literature is replete with proposed methods to cause the leading ends of MO marks to have the same shape as their trailing ends. All of these appear to proceed on the principle that by boosting the beam intensity at the leading end of the mark, it can be blunted to a shape that essentially constitutes a mirror image of the already blunt trailing end. This is accomplished by various stratagems, by which each data pulse is first processed to produce a final write laser drive pulse having increased amplitude at its leading edge, to yield a laser intensity boost at the leading end of the mark. Typically, this is accomplished by adding a ramp or step at the leading edge of each corresponding, essentially rectangular data pulse.
The Phase Change (xe2x80x9cPCxe2x80x9d) method of optical data recording is another means by which erasable data storage discs may be produced, typically in CD-RW (write/re-write) applications. Here, the disc contains a layer of substance whose structural phase (crystalline or amorphous) is changed by heat generated within its recording surface by a laser beam impinging on it. The phase state of a tiny region of the disc surfacexe2x80x94whether crystalline or amorphousxe2x80x94determines its optical characteristics, which are read optically to retrieve each datum thus recorded. The phase change method is fairly adiabatic, and, as in the MO method, excess heat is normally carried away by layers sandwiching the medium in which the data marks are created.
While PC, like MO, is a thermal optical data recording method, PC differs markedly from the MO method in that (1) in PC, it is the light beam, itself, that writes and erases the data marks, and (2) PC involves two thermal thresholds. There is one threshold for the xe2x80x9cerasexe2x80x9d mode and another, higher one, for the xe2x80x9cwritexe2x80x9d mode. These thresholds cannot be directly measured, as perhaps theoretically possible in the case of the temperature at an MO recording surface. Rather, they must be determined on the basis of many factors, such as the intensity of the light beam impinging on the surface, the rotational speed of the disc, the chemical and physical parameters of the recording surface, ambient conditions and other factors.
To write a PC markxe2x80x94i.e., a single datum or set of data, depending on the data modulation method employedxe2x80x94into the medium, it is necessary to that tiny region sufficiently to cause the material in that region to exceed its write threshold temperature and thus to change from its initial crystalline phase into an amorphous phase. This is accomplished essentially by melting it and then allowing it to solidify quickly in the new amorphous (i.e., data) phase. To control the shape of the mark, the laser heat must be removed very quickly to prevent reformation of the crystal phase (which characterizes lands). This is normally accomplished by causing each laser drive pulse to comprise a rapid succession of narrow pulses, each resulting in very brief laser light impact on the disc surface as it revolves. This succession of brief, intermittent pulses within each data pulse collectively creates the elongated microscopic mark. The sandwiching layers serve to withdraw any residual heat.
To erase a mark, the elongated microscopic region of the data track constituting it must be returned to its crystalline state. This, too, is accomplished by heating the region quickly. However, in the erase mode the region is allowed to cool more slowly, so that the crystalline structure can develop. Once again, this is normally accomplished by a rapid succession of laser pulses, which, in the erase mode, have a lower power level than in the write mode because the erase threshold is lower than the write threshold. The relative configuration and spacing of the narrow pulses comprising the erase pulse may be different than those within the write pulse, and the literature suggests many pulse designs that hopefully accomplish each respective purpose.
The marks created by the PC method, like those of the MO method, extend along a data track, because the disc is rapidly rotating while the laser beam switches between xe2x80x9conxe2x80x9d and xe2x80x9coffxe2x80x9d conditions. The duration of the laser beam pulse (i.e., the total duration of the particular set of narrow pulses comprising a single collective pulse) determines the length of the resulting mark (amorphous state) or intervening land (crystalline state). In some applications, CD-RW among them, the unrecorded disc is provided with a spiral tracking groove extending around the disc.
Like MO, PC normally employs ordinary PP tracking. In those applications in which the unrecorded disc is provided with a spiral groovexe2x80x94similar to the tracking groove in MO discs, and formed by similar methods, as discussed abovexe2x80x94tracking consists in causing the read beam to follow the central axis of the spiral groove, as in the MO case. However, unlike MO (which utilizes polarization detection), PC data retrieval consists simply in optical amplitude detection of read beam light reflected from the disc surface. This is based on the fact that the amorphous and crystalline regions of the disc surface will display different reflectivities.
As in the case of MO, there are many published methods to promote geometric symmetry between the leading and trailing ends of the data marks formed in the recording medium, for accurate detection of mark/land transitionsxe2x80x94i.e., for reliable data retrieval. As in the case of the MO method, these appear exclusively to deal with broadening the leading end of the mark to match its already broad trailing end, by increasing the energy input to the recording medium at the leading edge of each write laser pulse. In the PC method, this is typically accomplished by decreasing the relative spacing of the narrow pulses comprising each write laser drive pulse or by increasing their individual duration, toward the leading portion of each such collective pulse.
There are also published methods to alter the shape of certain types of marks in specific media to convert them into three-dimensional bumps. Typical of these are U.S. Pat. No. 4,719,615; 4,852,075; and 4,912,696. These all name Feyrer, et al as inventors, and are assigned to Optical Data Inc. Since they are quite similar, they will be referred to, herein from time to time, as the xe2x80x9cFeyrer Patents.xe2x80x9d These patents teach a dual-threshold, dual-layered concept, in which the inner layerxe2x80x94i.e., the one closest to the optical recording structure substratexe2x80x94is composed of a material whose shape can be changed by the injection of heat, typically from a laser beam. In the write mode, when the higher write temperature threshold is exceeded, the material expands to form raised bumps; in the erase mode, when the lower erase threshold is exceeded, the material contracts back to its initially flat configuration. The outer layer, of an entirely different composition, is bonded to the inner layer, and its purpose is to cause the heat-induced inner-layer bulge to be preserved after data writing, so that it can later be read. The heat causing the inner-layer bulge simultaneously causes the outer layer to become elastic, which elasticity ceases when the dual layer cools at that particular location.
It has been seen that MO and PC are each thermal optical data recording methods to produce a succession of erasable/re-writable, two-dimensional marks and intervening lands along a track, or tracks, in their respective data recording surfaces. In the case of the Feyrer Patents, a dual-layer, three-dimensional PC concept is also taught. However, it can be appreciated, even from the foregoing brief discussion, that the respective methods proceed on entirely different sets of physical principles, requiring quite different xe2x80x9cwrite strategiesxe2x80x9d to optimize mark formation for accurate data retrieval. This can easily be verified by comparative reference to the published literature on either method, despite the apparent fact that many of the write strategies described in these references are merely theoretical or based on computer modeling only, rather than having been rigorously tested in real-world circumstances.
Photoresist (xe2x80x9cPRxe2x80x9d) digital optical data recording likewise generates a succession of marks and lands along a track on a surface of an optical recording disc. Unlike the MO and typical PC methods, however, PR is a method for producing three-dimensional marksxe2x80x94i.e., pits or raised bumpsxe2x80x94and/or tracking grooves in the disc surface. For convenience, the ensuing discussion will emphasize pit generation, recognizing that the only difference in the PR method for selective production of indented or raised features lies in the proper selection of the photoresist material and the respective developing chemical(s), which selection procedures are well known in the art.
However, it is important to note that because of its ability to produce pits, the PR optical data recording method, in obvious contrast to the MO and typical PC methods, can be, and commonly is, used to generate disc masters, from which great numbers of commercial daughter discsxe2x80x94replica compact discs (CDs) or digital versatile discs (DVDs)xe2x80x94are pressed. Similarly, the PR method may be utilized to generate tracking groove (or tracking ridge) disc masters, e.g., for MO and PC disc substrates and CD-R/DVD-R blanks. While the Feyrer Patent PC teachings might, at first glance, appear to be applicable to disc mastering, the complex bonding processes and the nature of the materials necessitated by the Feyrer dual layer concept would not be conducive to commercially-practical, accurate disc mastering and replication.
But by the same token, a PR disc cannot conveniently be erased and re-written, and PR is therefore inapplicable to the chief purposes of the MO and PC methods: erasable/re-writable data storage. This merely demonstrates, once again, the fundamental differences existing between the respective methods of optical data recording.
The PR recording method is essentially a photographic engraving process. The recording surface of the master disc comprises a thin (xcx9c0.10 to 0.12 micron), photosensitive polymer resin layer of substantially uniform composition, which has exposure characteristics virtually identical to those of photographic film emulsions. Accordingly, PR is a purely photochemical (i.e., optical) method of optical data recording, as opposed to the MO and PC methods, which are thermal in nature. In other words, in PR, it is not the quantity of heat instilled in a selected small portion of the disc surface that exposes the photoresist; it is merely the quantity of incident light that determines if sufficient exposure occurs to enable formation of the desired marks. A threshold quantity of light is required to effect initiation of exposure at the photoresist surface. The extent and depth to which the photoresist below the immediate surface is exposed depends on the intensity of the impinging light and the optical characteristics of the photoresist material, itself. Because of light absorption and scattering within the photoresist medium, the width of the exposure within the medium typically decreases as its depth increases. However, as a general principle it is accurate to say that increasing the incident intensity will tend to increase the depth of exposure within the photoresist medium.
Since PR is normally applied to pit/land disc mastering, further reference will be directed toward that particular application. In accordance with convention, the input data are subjected to EFM (eight-bit-to-fourteen-bit) modulation, in the case of CD mastering. Here, sequential binary input data are converted into a sequence of spaced rectangular pulses, each of whose durations is nT, where T is the EFM clock period, approximately 231 nanoseconds (billionths of a second), and n is an integer from 3 to 11. In the case of DVD mastering, xe2x80x9cEFM Plusxe2x80x9d modulation is employed. This differs from EFM modulation principally in that: (1) eight-to-sixteen bit modulation is employed, and (2) the integer n may be 3 to 11, or 14. Every EFM or EFM Plus coded data stream always contains pulses and intervening temporal spacing comprising all of the possible nT durations. Thus each data pulse and each xe2x80x9coffxe2x80x9d period intervening between successive pairs of them are of nT duration, where, in each data stream, all permitted values of n must be represented in both the pulses and the intervening periods. Other modulation schemes have been used or proposed, and further modulation methods will be employed in the future, especially as the data recording density on disc masters and replicated discs inevitably increases. However, it should not be difficult to generalize from this discussion to encompass any such ordinary engineering modifications.
The pulses (xe2x80x9conxe2x80x9d times) ultimately generate pits in the disc surface, while the xe2x80x9coffxe2x80x9d times result in the intervening lands. Because the entire data stream is coded, both the pits and lands contain independent data. As is well known in the art, for proper data retrieval the resultant duty cycle (i.e., the ratio of pit length to pit-plus-next-land length) to should be approximately 50%. However, the duty cycle is averaged over a number of successive pit-land sequences, in a manner set by convention. Therefore, the length of each pit and of each land within such a sequence represents a separate, independent datum, despite the fact that the average duty cycle within the sequence set by convention is preferably maintained at approximately 50%.
Whether EFM (CD) or EFM Plus (DVD) coding (or any other modulation scheme, such as xe2x80x9cTwo-to-Sevenxe2x80x9d) is applied, the purpose is to insure that the data pits in the resultant disc are mutually spaced in a programmed manner to facilitate tracking. In commercial CD and DVD players tracking would otherwise prove difficult if, for example, a particular sequence consisted of a binary xe2x80x9conexe2x80x9d followed by a lengthy sequence of binary xe2x80x9czeros.xe2x80x9d With appropriate modulation, the CD or DVD player logic can anticipate the next pit (or land) in one of a specific number of succeeding locations. Although this may work reasonably well, it will be shown below that tracking can still be difficult, and this fact forms one basis for the present invention.
In PR recording, an EFM (or EFM Plus) coded waveform results in modulation of the intensity of a focused spot of light (normally from a diode or gas laser) impinging on the recording surface of the revolving disc. By properly synchronizing the rotational speed of the disc with the radial position of the beam relative to the disc center, this produces a track of narrow (width generally less than xcx9c1.0 micron), elongated latent images produced when the light spot is xe2x80x9con,xe2x80x9d interspersed with unexposed lands. When the entire spiral track (or collection of concentric tracks) is xe2x80x9cexposedxe2x80x9d onto its surface, the master disc is xe2x80x9cdeveloped,xe2x80x9d as in the case of ordinary photographic film. Immediately upon completion of that step, an etching solution is introduced to dissolve and remove the exposed regions of resist (or the unexposed regions, depending upon whether a positive or negative resist is used), which creates a succession of narrow, three-dimensional, elongated microscopic pits and intervening lands. The length of each pit and the length of each land represents an independent datum corresponding to a particular data packet in the original EFM (or EFM Plus) signal. As in the case of MO and PC recording, these elongated pits are microscopic, and there are normally many millions of them in the spiral disc track (or in each of the concentric tracks if there is more than one). Thus, when viewed microscopically, the pits will appear as a linear sequence of narrow, elongated, straight depressions, with intervening lands, lying alongside another such sequence radially (i.e., transversely) displaced from it by the fixed track pitch.
In CD and DVD applications, the pit and land lengths will each correspond spatially (i.e., in length) to the temporal duration of the corresponding portion of the coded data, so that when the discxe2x80x94or a daughter disc replicated from it, in the case of a disc masterxe2x80x94is played, the resulting information output will match the original information. To insure that all pits corresponding to nT in the original data are the same length for each particular value of n, the rotational speed of the disc master must be continuously varied during the recording process, and the rotational speed of the final disc must be correspondingly varied during playback, as is the case in all CD and DVD recording apparatus and players. In other words, the rotational speed must be varied, so that the linear speed is constant at every location. Accordingly, the rotational speed will vary in inverse relationship to the radial distance from the disc axis.
This is in contrast to a major MO applicationxe2x80x94computer hard drivesxe2x80x94where CAV (constant angular velocity) is maintained. There, CAV insures rapid data acquisition and retrieval, because no time is expended in changing rotational speed when the particular radial position of the desired data is reached. Of course, CAV can be applied to other optical data recording methods. However, where data retrieval at extremely rapid hard drive acquisition rates is not essentialxe2x80x94for example, in most CD-ROM and DVD-ROM applicationsxe2x80x94the introduction of necessary decoding logic to account for different data mark lengths corresponding to identical data values, depending on radial distance, may not be justified.
Returning to the PR method, the thickness of the photosensitive data layer (deposited on the much thicker glass or polycarbonate substrate) is selected to be identical to the desired pit depth. Thus, when the photoresist is fully exposed (through its entire thickness), flat-bottomed pits will be produced, whose depth will be the same as the thickness of that layer, and whose sidewall-to-base junctions, at least, are angular. Failure to fully expose the photoresist (resulting in residual photoresist at the bottom of the pit) has been found generally to produce xe2x80x9cnoisyxe2x80x9d data output readings, because of inherent roughness in the etched photoresist layer and greater susceptibility to recording laser noise. These compromise detection accuracy, because pit and land playback signal amplitude is affected by surface characteristics.
Accordingly, the thickness of the photoresist layer and the exposure levelxe2x80x94and thus the resulting pit depthxe2x80x94are normally selected for optimal detection in a manner well known by those skilled in the relevant art. Pit width is determined by the power and effective width of the recording beam. The latter is determined by the wavelength of the write laser utilized and the numerical aperture (xe2x80x9cNAxe2x80x9d) of the focusing means, in a manner likewise well known in the art. The transverse (i.e., radial) sectional shape of the pit, whether rectangular or trapezoidal, may be controllable by the optical characteristics of the photoresist material and by the particular focusing configuration selected, as described in the Dil and Sugaya, et al patents referred to below. Finally, the length of each resulting pit will be primarily determined by the duration of the corresponding EFM (or EFM Plus) data pulse, as will be the length of each intervening land.
Once recorded, a PR-generated disc master is converted to a metal stamper by conventional galvanic processes, and daughter discs are pressed from it. If sufficient skill and care are exercised, the stamper will be a virtually exact mirror image of the master, and the replicated discs will likewise be virtually exact copies of the master. Less stringent controls can yield a stamper that is a reasonably good mirror image of the master, but replicated discs whose reproduced pits do not exactly reproduce the cross-sectional shape of those in the master. The latter is typically the result of imprecise molding methods, yielding replicated disc pits whose cross-sectional shapes in the radial direction display rounded corners, rather than the normally crisply angular corners of the master disc pits.
Double-sided discs can be generated by utilizing two moldsxe2x80x94one for each side. Furthermore, double-layered discs may be made by layering two data bearing surfaces on each side of the resulting CD or DVD, each layer again created by a separate stamper produced from its disc master. Retrieving the data from each layer depends on conventional means to detect reflection of light partially passing through the interface between the layers. Theoretically, multi-layered, two-sided CDs and DVDs (and anticipated future expansions beyond present DVD-ROM data densities) could thus be produced, all based on the same PR disc-mastering procedures.
The process for generating tracking groove disc masters (e.g., for MO and PC disc substrates and CD-R/DVD-R blanks) by the PR method is quite similar to that employed for pit production. Indeed, the groove mastering process is simplified by the fact that the desired spiral groove is normally continuous over all or most of the disc master. Thus, a tracking groove master can be generated merely be providing a constant amplitude input to the writing light source and synchronizing the radial position of the write beam with rotational speed of the disc and timing, by conventional methods. If a discontinuous tracking groove master is desired, this merely requires selective write beam extinguishments and radial position adjustment, likewise by conventional methods. By ordinary engineering modifications within the basic principles, virtually any sort of tracking groove master may be generated by the PR method. Likewise, by a conventional selection of photoresist material(s) and development chemical(s), a tracking ridge master may likewise be produced for any application requiring it.
One common tracking groove configuration is dictated by the specifications for unrecorded CD-Rs and DVD-Rs, which require xe2x80x9cwobbledxe2x80x9d grooves. In these applications, the tracking grooves in the unrecorded discs are not merely spiral. Rather, a radially sinusoidal displacement is superimposed onto the spiral tracking groove as it is created. In CD-R grooves, the specified amplitude of the sinusoidal xe2x80x9cwobble,xe2x80x9d in respect to the longitudinal axis of the xe2x80x9cunwobbledxe2x80x9d spiral groove, is xc2x130 nanometers (nominal), and its frequency is approximately 22 kHz, at nominal linear velocity of recording. A separate input to the write beam radial positioning means creates this selective wobble.
Regardless of the particular application, the PR method is essentially an etching process, and a certain amount of roughness occurs on the pit sidewall surfaces is thus inevitable. While this has not proved to be a particularly significant problem in CD mastering applications, where data retrieval is essentially diffraction-interference based, the PR method of disc mastering appears not entirely conducive to production of DVD masters from which commercial DVD-ROM""s can be rapidly manufactured with a low rejection rate. This is because DVD data retrieval is a necessarily more sophisticated process, relying on phase comparisons, as well as the basic diffraction-interference effects. The inherent roughness of PR-generated data pits impedes accurate DVD data retrieval.
Furthermore, this problem of roughness in PR-generated disc masters can only become more troublesome as data densities increase beyond the present DVD-ROM level of approximately 4.2 gigabytes per data layer, and data retrieval strategies necessarily become more sophisticated. Few would argue that a much higher data density than can presently be provided even by two-sided, double-layered DVDs would not be highly desirable, and this will undoubtedly occur as practical ultra-violet and perhaps higher-frequency lasers and other beam sources are developed, and narrower (and correspondingly shorter) pits and track pitches result. When this occurs, the inherent roughness of PR-generated disc masters may further limit their utility.
The characteristic roughness, and the fact that PR-generated pits are normally rather steep-sided and display angular corners, also result in certain difficulties in separation of daughter discs from the stampers by which they are reproduced. Pits that are trapezoidal in cross-sectionxe2x80x94as described, for example, in U.S. Pat. No. 4,209,804 (Dil), U.S. Pat. Nos. 4,230,915 and 5,459,712 (Sugaya, et al), in contexts unrelated to separation of daughter discs from stampersxe2x80x94might alleviate that problem, if pits of the shapes described therein can actually be generated by the PR method. This is theoretically possible, as briefly discussed above, although it is by no means certain that the various parameters can be selected and controlled, in practice, to achieve the desired results. However, even if such pit configurations are possible with the PR method, this inherent pit roughness and angularity problems can, at best, only be somewhat alleviated, even by the methods described below in the context of the present invention. In all probability, it may not be possible to completely eliminate them.
By contrast with the MO and PC methods of optical data recording, heat is not directly involved in PR pit production; it is merely the time-integrated light power density acquired from the focused recording light spot that exposes the master disc. The thin photoresist layer absorbs very little heat, since its thermal absorption is quite low and only a small fraction of the incident light photonsxe2x80x94on the order of 1 in a millionxe2x80x94are captured by the photosensitive component of the surface material. The PR recording method is, therefore, an adiabatic process. Thus, PR recording write strategies are typically much simpler than those that would be optimal for the otherxe2x80x94thermalxe2x80x94optical data recording methods. Indeed, in PR CD applications a binary pulse-length modulated waveform (such as the xe2x80x9crawxe2x80x9d EFM or EFM Plus waveform) that has only minor adjustment to the lengths of its xe2x80x9conxe2x80x9d pulses may often be directly applied to modulate the laser beam to effect reasonably high quality, full-depth pit recording in a thin photoresist layer.
Dye-polymer optical data recording is addressed in U.S. Pat. No. 5,297,129 (hereinafter, xe2x80x9cthe ""129 Patentxe2x80x9d), assigned to the assignee of the present application. Like PR, it is presently employed to reproduce EFM and EFM Plus coded digital or digitized data streams as tracks of three-dimensional, elongated microscopic pits and intervening lands in optical data storage structures, and is likewise utilized to produce disc masters from which commercial CD-ROMs and DVD-ROMs are stamped. It is also utilized to produce tracking groove masters for MO, PC, CD-R and DVD-R production, and to produce masters for xe2x80x9chybrid CDsxe2x80x9d (incorporating an annular CD-ROM portion containing recorded data, plus an annular CD-R portion containing a tracking groove in an optically recordable medium on which further data can later be recorded).
However, unlike PR, dye-polymer recording is a thermal process, and thus it proceeds on the basis of physical principles quite different than those underlying PR. Likewise, it can easily be seen that dye-polymer recording, although a thermal process, differs from MO and PC, in that the physics of these three processes are fundamentally different.
Essentially, the dye-polymer process, applied to disc structures as the principal example, comprises selective expulsion of the photo-thermally active recording layer of a disc (or disc master, as may be the case). This surface is comprised of a mixture of a polymer (nitrocellulose having been commonly employed) and a dye whose color is complementary to that of the write beam (normally from a laser source) to promote maximum heat absorption. If, for example, the write beam consists of a red or infrared diode laser, the corresponding dye color would likely be a suitable shade of blue or blue-green. On the other hand, if an argon ion laser write beam (or other blue light source) were utilized, a red dye would probably be chosen.
Theoretically, if a sufficiently powerful energy-transferring beam were employed, the dye content could be reduced or perhaps entirely eliminated. But the result would otherwise be identical to the normal dye-polymer process, and these xe2x80x9clow-dyexe2x80x9d and xe2x80x9cdye-lessxe2x80x9d methods would thus merely be species of the general dye-polymer method and included within it for the purposes of the present invention. But unlike the complex, dual-composition data layer taught in the Feyrer Patents, for example, the mixture comprising the data layer in any such dye-polymer process would be substantially uniform in composition throughout that layer.
For clarity in presentation, primary reference will once again be made to disc embodiments and applications. However, it will be understood that the principles underlying dye-polymer optical data recording apply equally to other recording-structures, such as cylinders, cards and data wafers, as does the present invention.
In the case of discs, the (photo-thermally) active layer is spin coated onto a surface of a supporting disc substrate, normally constructed of glass or polycarbonate plastic, or from another material of suitable strength and optical and thermal properties. In the case of xe2x80x9cfirst surface recording,xe2x80x9d where the write laser is directed externally xe2x80x9cfrom the topxe2x80x9d onto the active layer (e.g., in some mastering applications), the substrate need only possess suitable strength and thermal expansion properties, compatibility with the material comprising the active layer and compatibility with the substances utilized in the subsequent galvanic processes. However, in the more typical xe2x80x9csecond surface recordingxe2x80x9d case (e.g., CD-R, DVD-R and some disc mastering applications), where the write laser beam is directed xe2x80x9cfrom the bottom,xe2x80x9d through the substrate and focused onto the active layer lying at or near the opposite surface of the substrate, the substrate material must also possess a suitable level of transparency and index of refraction, in respect to the particular write beam wavelength. To decrease reflectivity from the xe2x80x9cbottomxe2x80x9d surface of the substrate, an appropriate anti-reflective coating may be applied there. Whether first or second surface recording is performed, the detailed selection and implementation processes are well within the capacity of an optical engineer with ordinary skill in the art, based on the discussion herein and in the ""129 patent, and on the nature of the physical processes involved.
During pit formation (and also during data retrieval) the disc spins rapidly. The speed of rotation varies continuously, in inverse ratio to the distance of the write beam from the center of the disc, to insure that the linear speed of pit formation (i.e., the speed of the relative longitudinal movement between the write beam and the track) remains constant. As in the case of PR recording, this is to insure that each pit spatially representing a data packet of a given temporal duration, nT, is of the same length as all other pits representing that same nT duration, to insure that upon playback (where angular speed is correspondingly varied to insure constant linear speed) all pits of the same length will yield the same output value. CAV recording and playback could be utilized. However, this is not generally done, for the reasons briefly explained above.
Expulsion of material, in the context of dye-polymer optical data recording, is believed to comprise some combination of decomposition (breaking of polymer chemical bonds to form smaller molecules); explosion (forceful discharge of material by chemical reaction); fluid flow (plasticization or melting), where heat-induced expansion causes flow upward and out of the formed pit); and partial compression of material from altered surface tension.
These effects, and the combination present in any expulsion context, will probably vary to some extent depending on the specific dye-polymer recording configuration. For example, in first surface disc mastering applications, an open pit is created when material is expelled and/or flows upward. In second surface disc mastering operations, where the reflective layer is normally above the active layer (i.e., on the extreme opposite side from the initial point of impact of the write beam onto the substrate), some re-condensation of material held by the reflective layer may occur as some of the latter collapses into the formed pit. Finally, in CD-R-/DVD-R (second surface) applications, the xe2x80x9cpitxe2x80x9d is actually a bubble-like void formed by gasification, deformation and refractive index change between the reflective layer and the substrate, as the active layer material within it is expelled. That void displays distinct optical characteristics that can be read xe2x80x9cfrom the bottomxe2x80x9d in the same manner as the pits (viewed, from the bottom, as bumps) of a CD-ROM are read, which is why CD-Rs and DVD-Rs can be read in ordinary CD and DVD playback devices.
In any event, expulsion begins to occur when the material in a tiny region of the active layer exceeds the thermal threshold of the moving medium. Somewhat as in the case of PC recording, this thermal threshold is not readily measurable, nor is it directly proportional to the instantaneous intensity of the write beam causing it. Rather, it is a temperature level induced in a minute volume of material by a quantity of heat absorption in that tiny portion of the medium, as it moves relative to the energy-transferring beam creating it. Thus, it depends on many factors, including (but not limited to) the temporal and spatial write beam intensity profile, the thickness of the active layer, the instantaneous speed of rotation of the disc, the exact nature of the materials comprising the dye-polymer material and their proportions, the nature of the disc substrate and ambient conditions. All of these can vary from batch to batch and day to day. An analogy might be made to a pot of water on a stove. Here, boiling results as the energy induced into individual water molecules causes them to undergo a change of physical state from liquid to gaseous, at a temperature threshold that is reached at that molecular level. The amount of heat applied is only one factor in determining when and how rapidly the water will boil. Other factors include the quantity and purity of the water, the nature and thickness of the water container and ambient conditions.
Accordingly, write strategies applicable to the dye-polymer method of optical data recording must be based on the physical principles and peculiarities of the dye-polymer recording process (as opposed to the very different MO and PC processes, for example), and must provide the flexibility to facilitate optimization for each set of blank discs and recording conditions.
To provide the context for a discussion of particular write strategies, and particularly to those that form the basis of the present invention, the dye-polymer method will first be explored, with emphasis on first surface disc mastering applications, merely for clarity.
A laser output typically generates the tightly focused write beam, although neither this discussion nor the present invention is limited to the use of a laser write beam. Ion beams, electron beams and many other intense optical or pseudo-optical energy-transferring beams may be used. In the case of a solid-state (e.g., a diode) laser, a digital data stream (normally, a series of spaced rectangular pulses, resulting from EFM or EFM Plus coding of binary data) can form the drive input to the laser, thus directly controlling its output. This is because a solid-state laser will respond essentially instantaneously to that input. However, if a gas laser is employed, external modulation must normally be employed, since gas lasers typically cannot react to input fluctuations as rapidly as can solid state lasers. Accordingly, in the case of a gas laser, control of the laser output is typically accomplished by some means interposed in the path of the emitted beam, such as an acousto-optic modulator (xe2x80x9cAOMxe2x80x9d), which is controlled by the digital data stream input, or by a selective derivative of that input. But there is no essential difference between these two methods (or, indeed, between either of them and a method utilizing some other intense energy-transferring beam), because in either case it is the amplitude waveform of the modulated digital data stream that ultimately determines the intensity profile of the beam that impinges on the active disc layer. The choice is dictated by the particular laser (or other beam) type chosen, in a manner well within the ability of an ordinary practitioner, in light of the teachings herein.
The impinging write beam has typically displayed an essentially circular cross-section with an approximately Gaussian (bell-like) intensity distribution measured diametrically across the circle. In other words, the cross-section is essentially an Airy Disc. The diameter of a concentric circle that will enclose 50% of the total light power in this Airy Disc spot is less than 1 micron in the usual case of a laser write beam of typical frequency. When the focused spot is turned xe2x80x9con,xe2x80x9d a significant fraction of its light power is absorbed in the thin dye-polymer layer, causing heat to be generated in the tiny area illuminated. Almost immediately after it is deposited, this heat will begin to diffuse away from the minute area where it was originally injected. If enough heat is coupled into a small volume (typically less than 1 cubic micron) of dye-polymer material for a sufficiently long period of time (xcx9ca few tens of nanoseconds or less), the thermal threshold of the moving material in that small volume will be exceeded and expulsion will occur.
Particularly in the case of CD/DVD mastering applications a portion of the material that melts or plasticizes and is caused to flow out of the formed pit will re-solidify at the cooler top of the pit. This will create an elevated lip or xe2x80x9cbermxe2x80x9d surrounding the pit that is formed when the material originally occupying that space was expelled. Because the disc is spinning during this process, the resulting bermed pit will be elongated, and will continue to elongate as long as the thermal threshold at the trailing end of the forming pit is exceeded.
At CD or DVD recording speeds pit writing in the thin dye-polymer layer is not a truly adiabatic processxe2x80x94i.e., a portion of the heat created in the thin optically active layer may migrate (diffuse) some distance from the point at which it was originally generated, in part as a result of the complex expulsion process. This xe2x80x9cthermal smearingxe2x80x9d affects the size and shape of the recorded pit. Accordingly, in dye-polymer CD/DVD recording, the applied write strategy must precisely manage the generation and resulting flow of heat within the thin active layer to insure that the resulting stream of recorded pits and lands may be accurately tracked and read. Clearly, an optimal write strategy will depend on the physical, rheological and optical parameters of the active layer and substrate, the shape and power range of the focused recording spot, the range of velocities at which the disc is rotated, the range of nT pulse and spacing durations in the recorded data stream (i.e., whether in the context of CD or DVD recording or some advanced, higher-density write strategy) and the desired characteristics and tolerances of the playback signal that will be obtained when the final stream of pits is ultimately tracked and read.
Clearly, as in the case of PR mastering, for accurate data retrieval (i.e., accurate determination of all individual pit and intervening land lengths) by a CD or DVD player, the write strategy should optimize detection of the pit/land transitions. One way of improving the accuracy of pit/land detection is to cause the pits in each track of the master disc to display three-dimensional geometric symmetry (i.e., to cause the shape of the two ends of each pit to mirror one another). Another way to improve detection of pit/land transitions, particularly in discs replicated from PR or dye-polymer masters, is to employ appropriate pre-compensation to adjust the duty cycle and/or depth of the three-dimensional marks recorded on the disc master, based upon their respective radial positions on the disc. The latter are discussed, respectively, in U.S. Pat. Nos. 5,608,711 and 5,608,712 (hereinafter, xe2x80x9cthe ""711xe2x80x9d and xe2x80x9cthe ""712xe2x80x9d patents), assigned to the present assignee. These methods, which will be discussed below in various contexts, can be applied individually or together.
Accurate data retrieval also requires accurate tracking. So the write strategy must additionally provide sufficiently precise pit shapes and land configurations to insure that even relatively inexpensive CD and DVD playback devices can accurately follow the data track(s), while they perform accurate data retrieval (detection of pit/land transitions and, thus, individual pit and land lengths and corresponding data values).
Unfortunately, this is complicated by the fact that the criteria inherent in accurate data retrieval (high frequency, xe2x80x9cHFxe2x80x9d) detection and in accurate push-pull (xe2x80x9cPPxe2x80x9d) tracking, required in all pre-recorded CD applications, are mutually exclusive. Similarly, there exists a fundamental tradeoff in PP and groove reflectivity, in unrecorded CD-R and DVD-R specifications. Hybrid CD specifications require analogous compromise. Indeed, it is safe to say that in nearly all optical data recording methods, some compromise must be made between two or more detection requirements.
When writing a pit by the dye-polymer method, the write beam is focused to cause its diameter (i.e., the diametric distance from a xc2xd power point to the opposite xc2xd power point in the Gaussian distribution of power within its cross-sectional Airy Disc) to be approximately the width of the pit to be created (conventionally measured halfway between the disc surface and the base of the pit). The read beam diameter is generally double the width of the pit.
According to specification, the CD track pitch (xe2x80x9cTPxe2x80x9d) is between 1.5 and 1.7 microns, the nominal value being 1.6 microns. The length of an EFM-coded CD pit is nominally 0.3 micron per T, where the pit spatially represents an input data pulse of nT temporal duration. The width of a CD pit (again, measured at half depth) and the diameter of the write beam creating it are each approximately 0.5 micron, i.e., approximately TP/3. On the other hand, the read beam is approximately double that width, or about 1 micron wide. Since various laser wavelengths are utilized in CD recording, the numerical aperture of the objective lens focusing the beam must be selected to yield a beam spot of the same diameter regardless of the beam source, so that the pits will be the same width regardless of the apparatus used, to insure that the resulting pits may be uniformly read. The spot diameter, d, is determined by the formula d≈0.5xcex/NA, where xcex is the beam wavelength in vacuo, NA is the numerical aperture and d is the diameter of the resulting spot. In the case of CD playback, for example, xcex=0.780 micron, and NA=0.45, so d≈0.9 micron.
Similar proportions apply in DVD applications, although the dimensions pertaining to DVD recording and reading are approximately 50% of those pertaining to CD applications, reflecting the correspondingly shorter channel bit lengths of DVD marks. Presumably, future higher-density applicationsxe2x80x94utilizing higher frequency (i.e., shorter effective wavelength) write and read beams, smaller pits and narrower track pitchesxe2x80x94will employ similar relative proportions.
Most dye-polymer disc mastering systems utilize DRAW means, whereby the readability of the formed pits can be determined in real time as they are created. These generally utilize a monitoring beam, whose light reflected from the disc is detected and analyzed, as described, for example, in U.S. Pat. Nos. 4,809,022 and 4,963,901. Since DRAW mastering is now reasonably familiar to those skilled in the relevant art, it is believed unnecessary to further discuss it here.
The same read beam can be used for HF and PP detection, although it is suitably divided, as described herein and as well known in the art. However, although both are based on measurement of light reflected from the disc, HF and PP detection proceed in accordance with opposing principles.
In HF detection, it is the reflectivity contrast between pit and land portions of the data track that is observed. The goal in this regard is to cause pit areas to be seen as very dark and land areas to be seen as very bright. This is so that when the beam passes through a pit/land transition (as the disc revolves rapidly and the track of pits and lands moves at correspondingly rapid linear speed relative to the read beam), the established level of detected brightness constituting a transition event will be reached and passed clearly and quickly. If the pit/land transition events can be detected with great accuracy, this will result in precise determination of the corresponding pit and land lengths, from which the original information can be reliably regenerated.
The desired HF optimization is achieved with an effective pit depth (noting that each pit will normally posses a curved base, caused by the dye-polymer expulsion processes) equal to xcex/4, where xcex here is the wavelength of the (typically, laser) read beam within the substrate material (since discs are normally read from the second surface). This will create a xcfx80 (180xc2x0) phase shift in the reflected light, effectively canceling out, by interference, the small proportion of incident light not already scattered away by diffraction. By contrast nearly 100% of incident light is reflected from the essentially flat land areas. It can easily be seen, then, that with xcex/4 pit depth the change in reflected light detected at each pit/land transition will be very abrupt, thus facilitating accurate detection of pit and land lengthxe2x80x94i.e., accurate HF detection.
By contrast, PP detection generally measures the quantity of light diffracted from the pit at an angle in respect to the perpendicular direction. This is normalized with the known or observed reflectivity of the disc surface, to provide comparative values in the particular context. Thus, in CD applications, radial PP detection is merely an amplitude comparison of detected light on either side of the longitudinal track axis (whether within a pit or a land area). When more reflected light is received on one side of the PP detector than the other, the PP servo moves the read beam in the opposite direction, radially, until detection in the two halves is equalized, indicating proper tracking. An effective xcex/8 groove depth, producing a xcfx80/2 phase shift, optimizes PP detection, rather than the xcex/4 groove depth and corresponding xcfx80 phase shift that optimize HF detection. As mentioned above, a similar dichotomy in criteria exists between PP and unrecorded groove reflectivity determination in CD-R and DVD-R applications, and other optical recording applications, such as hybrid CDs, present analogous dichotomies.
To promote reflection, and to increase playback amplitude (without correspondingly increasing reflectivity contrast), the entire disc is usually provided with a thin coating of aluminum (or other suitable highly reflective, easily applied material). Both detection procedures are based on the fact that, in the context of the dimensions inherent in the process, the pit acts as a single-slit diffraction grating, scattering most of the incident light in a direction radial to the disc and returning only a small portion back in the direction of the incident beam for detection. But the two processes are inherently opposed.
Because of these well-recognized problems, the concept of xe2x80x9cFigure of Meritxe2x80x9d has become widely applied in connection with development of write strategies for all optical data recording, including tracking groove mastering (although principal reference will here be made to PR and dye-polymer data disc mastering). Simply stated, the Figure of Merit is a weighted function that measures overall conformity to the established standards. In the case of pre-recorded CDs and CD masters, its contributing factors are the amplitude of HF detection, the amplitude of PP detection, minimization of cross talk between radially adjacent portions of the recorded data track and other factors not as relevant to this discussion. In short, the ultimate purpose of write strategy development in optical data recordingxe2x80x94particularly disc mastering of any nature by the PR and dye-polymer methodsxe2x80x94is to maximize the Figure of Merit.
The invention described and claimed in the ""129 patent constitutes a major step toward this goal. By its specific reference to a thermal optical data recording method, it excludes the PR method (which is a purely optical, not a thermal data recording method), as well as MO and PC (which do not produce marks of the types identified in the ""129 patent). Thus the principal context of the ""129 patent is the dye-polymer method. Because it is a thermal process, involving plasticization (softening or melting) and fluid flow to a certain extent, the dye-polymer method, by its very nature, tends to produce pits with smoother surfaces than those generated by the PR method. As discussed above, this facilitates data retrieval (particularly in DVD and other ultra-high data density applications), as well as separation of injection-molded replicas from their stampers.
Focusing, then, on the dye-polymer method, particularly in reference to disc mastering, the ""129 patent provides a superior method and apparatus for generating pits for improved HF detection. The published prior art write strategies (relating, perhaps exclusively, to the MO and PC methods) apparently hope to deal with geometric asymmetry in data marks by increasing the write beam power at the leading edge of write pulses to blunt the leading end of data marks (which are otherwise tapered because of time delay in heat buildup in the moving medium at the leading ends of the marks), to match the blunt trailing ends caused by abrupt shutoff of write power pulses at their trailing edges. By contrast, the essence of the ""129 patent invention is to modify the trailing edge of the write pulses to taper the trailing ends of the pits to match their already tapered leading ends.
In the context of the ""129 patent, the term xe2x80x9ctaperxe2x80x9d refers generally to the elongated pit leading end broadening caused by initial heat buildup as pit formation begins, or to the elongated trailing end narrowing brought about by the moderated pulse trailing edge decline in write power taught in the ""129 patent. That is to say, the term is not limited to leading and trailing end shapes that are pointed, and may include elliptical or other elongated (as opposed to bluntly semi-circular) shapes, the important consideration being that whatever the shape of the leading end elongation, it is an important purpose of the ""129 patent to provide a trailing end elongation to mirror it. It is in that broader sense that the term xe2x80x9ctaperxe2x80x9d is used in the context of the present discussion and invention, as well.
Fundamentally, the ""129 patent describes and claims the concept of a trailing region of the write pulse within which the write beam power is decreased over time, rather than abruptly shut off. While the preferred embodiment describes a ramped trailing edge, it can be seen by those with ordinary skill in the relevant art, having a basic understanding of the physical principles involved, that the trailing edge modification claimed in the ""129 patent may, for example, alternatively comprise an exponential decay, a series of steps or even a single intermediate step, all equivalently within the claimed concept. As described in the patent, this is because the trailing end taper is generated by the moderated decrease in write power, causing a slower drop through the thermal threshold of the moving medium than in the case of an abrupt power shutoff.
As described in the ""129 patent, this not only improves HF detection, but it also facilitates optimization of the write strategy to account for fine differences in blank disc batches, ambient conditions and other factors. This is because the trailing region parametersxe2x80x94whatever its chosen power decline profilexe2x80x94can more easily be adjusted for optimal mark configuration than in the case of an abrupt terminal power shutoff, where the only easily adjustable parameters might be write pulse length, write level and the shape and/or extent of a leading edge power boost (if any).
The ""711 and ""712 patents provide additional improvements in HF detection. Specifically, the ""711 patent teaches methods and apparatus for adjustment of the data track duty cycle in certain areas of the disc master to compensate for effects of the replication process. Similarly, the ""712 patent provides a method and apparatus for selectively adjusting pit depth in disc masters to improve HF detection in replicated discs. These strategies may be applied individually, together and/or with the teachings of the ""129 patent, according to need.
Yet despite these significant steps toward Figure of Merit maximization, the ""129, ""711 and ""712 patents do not emphasize another of its major componentsxe2x80x94tracking accuracy.
A recent European patentxe2x80x94EP 0 837 454 A2 (hereinafter, xe2x80x9cSchoofsxe2x80x9d)xe2x80x94deals, to some extent, with the problem of increasing the Figure of Merit by improving PP detection in an optical data disc (apparently, a dye-polymer recording), without unduly compromising HF detection. The proposed solution is to increase the intensity of the write beam between write pulses to a level just above the thermal threshold of the moving medium. This creates a narrow, shallow groove in the land area intervening between successive pits, which essentially increases PP tracking signal strength between pits with hopefully little negative effect on HF (i.e., pit/land transition) detection accuracy.
To a certain degree, the method taught there does satisfy the PP optimization criteria, in that the land groove can be made to be approximately xcex/8 in depth. However, because this is accomplished by reducing the write beam intensity to near the thermal threshold, the resulting land groove must necessarily be quite narrow. But this actually compromises PP detection, because optimal PP detection is realized with a groove that is wider than one that would optimize HF detection. Furthermore, HF detection is not significantly addressed by Schoofs. Indeed, the logical extension of the Schoofs teachings would be to further increase beam intensity between pits to widen the groove for better tracking. But that would compromise PP detection by deepening the groove and would also compromise HF detection by causing pit/land transitions to be more difficult to detect, thus negatively counterbalancing the proposed PP improvements in the overall Figure of Merit.
The above-mentioned Dil patents (U.S. Pat. Nos. 4,209,804 and 4,230,915) address tracking of the pits themselves, to some extent. Both of these patents discuss PP detection and the problems associated with it. The first of them seeks to alleviate such tracking problems by sloping the sides of the pits to cause the pits to appear to be shallower than they actually are, for the purposes of PP detection, while allowing the HF detector to see an effectively greater depth. The second modifies the first by proposing that the index of refraction of the medium in which the pits are ultimately created be chosen, together with the side angles of the pits, to optimize that result. The goal of these teachings is, of course, to attempt to optimize the Figure of Merit.
But there are several fundamental problems with either of those teachingsxe2x80x94which are both based on the PR optical data recording method: (1) it is difficult to consistently produce slope-sided pits by the PR method, which is essentially an etching process and which, therefore, tends to generate pits whose sides are perpendicular to the disc surface; (2) slope-sided, PR-generated pits will expose more of their characteristic roughness (i.e., noisiness) to the HF detector, which makes DVD and ultra-high density data retrieval difficult; yet (3) pits whose sides are perpendicular to the disc surface increase the previously discussed problem of difficult stamper-replica separation.
Neither of these latter two patents nor the above Schoofs patent addresses the problem created by the fact that the optimal groove/pit width for PP detection is not the same as for optimal HF detection.
U.S. Pat. No. 5,459,712 (Sugaya, et al) also discusses certain problems related to pit shape optimization in a general, theoretical sense, in the context of increasing CD data density. However, this patent, like the other references, merely proposes a different compromise between the conflicting requirements for optimal PP and HF detection.
Therefore there is a need for a method, an apparatus and resulting pit and land geometry by which the Figure of Merit in optical data recording discs, disc masters and other optical recording structures displaying three-dimensional marks can be reliably improved by optimizing tracking detection along the entire track of data marks without substantially compromising HF detection and other specifications.
Furthermore, there remains a need, in optical recording structures, for data features of improved cross-section, so that when a data track is read, cross talk between the data features on that track and the adjacent track(s) is reduced, thus likewise improving Figure of Merit.
There also remains a need for a method, apparatus and resulting groove geometry by which tracking can be improved in CD-R, DVD-R, MO, PC and other pre-groove optical data recording structures.
Additionally, there remains a need for improved data feature configurations in optical recording structure mastering applications, so that replicated discs may be separated from a stamper with reduced replication error resulting from material clinging to unwanted crevices in the stamper surface, and to consequently improve repeatability in the replication process.
This goal of this invention is to improve Figure of Merit in optical recording structures, principally by providing three-dimensional features with improved transverse cross-section. These features may include data marks, such as pits and bumps; three-dimensional land areas; and tracking grooves and ridges. As can readily be seen from the foregoing examples, these features can be indented into the data layer (e.g., data pits) or raised above it (e.g., tracking ridges). As applied to creation of data features, the invention also improves tangential (i.e., along the data track) cross-section, to provide clearer pit-land transitions, as compared to Schoofs and other prior art teachings.
In the principal embodiments, these improved features are characterized, in part, by the fact that their transverse cross-sectional shape may display a continuous slope from one transverse end to the other transverse end. In the context of the present invention, xe2x80x9ccontinuous slopexe2x80x9d means a curvilinear shape of perhaps various curvatures, but substantially lacking angular junctions between any adjacent portions. For example, a sinusoidal waveform, or a complex waveform representing several sinusoidal waveforms summed together, will display a continuous slope by this definition. By contrast, a pit or groove with a trapezoidal section would not display a continuous slope, as the junctions between the base and sidewall would be angular, rather than curved. On the other hand, a section displaying even somewhat rounded corners would display a continuous slope.
Another element of the inventive concept, in the principal embodiments, comprises berm reduction or elimination. In this regard, berm reduction is shown by reduction in the ratio of maximum vertical displacement of the berm, from a defined index level, to maximum vertical displacement of the principal feature, from that index level. The index level may or may not be at the surface level of the disc, because the disc may contain multiple, vertically-spaced data layers, and/or the feature may be defined in respect to a trough or mound whose level is different than the surface level of the disc or of the particular data layer level.
If, for example, the feature were a pit (or tracking groove) in the structure, then the ratio would be the maximum height of the berm, measured upward, divided by the maximum depth of the pit or groove, measured downward: a first, relatively small value according to the invention. In both cases, the measurement would be made from the index level, which is the level from which the berm begins its ascent, on either side.
In the case of a data bump or tracking ridge (found typically in a structure pressed or otherwise mirrored from the foregoing structure, or perhaps created by essentially inverting the waveform controlling the write beam, as will be discussed below), the ratio would be the height of the feature divided by the depth of the berm-analogous indentation: a second, relatively large value according to the invention. That second value approximates the reciprocal of the first value.
Between those first and second values is a range of excluded ratios that might be realized without utilizing the improvements disclosed herein. These, of course, reside in the prior art.
In certain embodiments, the feature is multiple. That is to say, for example, that an improved pit might reside within a trough imposed in the data layer. Likewise, a data bump might be raised above a mound, likewise rising in the data layer. In either of those cases, the defined index level, in respect to which the transverse cross-section of the feature is analyzed, would be the level of that secondary feature from which the principal feature begins its vertical departure. For example, in the case of a pit within a trough, where berms rise from either side of the pit, the index level would be that level from which either berm begins its rise from the trough, wherever that occurs (see, e.g., FIG. 7).
A further element of the inventive concept, in the preferred embodiments, comprises the narrowing of berms, whether or not their height is reduced. Such berm narrowing reduces cross talk between radially adjacent track portions, particularly in high-density (i.e., small track pitch) applications, and also reduces the effect that the writing of surface features (e.g., pits or a groove) in one portion of the track might otherwise have on the radially adjacent portions.
Various apparatus are disclosed, by which the claimed structure displaying features of improved transverse cross-section may be produced.
In the preferred embodiments, this is accomplished by providing a secondary write beam, directed more widely and with less intensity than the primary write beam, but whose focal point is normally co-axial therewith, to create a shallow trough along selective portions of the data track, within which the individual data pits will reside. By selectively adjusting the intensity and focus of these primary and secondary beams, the secondary beam may be caused to reduce the pit berm height (or eliminate berms entirely) by expelling some material surrounding the pits. Furthermore, the secondary beam may be made to cause the pits, of optimal sectional dimensions for HF detection, to be formed within a selectively continuous trough of optimal sectional dimensions for PP detection. The secondary beam may be continued into the land areas or, if desired, maintained in an xe2x80x9conxe2x80x9d state, to create an optimally-configured tracking trough along all or any selective portion of the data track. By selectively adjusting the depth of the resulting trough, HF detection will not be compromised by the improvements provided for improved PP detection, because the pit/land transitions relative to the base of the formed trough will be as easily detected as if no trough were implemented.
In further embodiments, the essential dual-beam concept of this invention may be applied to provide similar improvements in other optical recording structures. For example, the secondary beam may be utilized to provide a relatively wide, shallow trough in PR applications, by generating a secondary, lesser exposure of the photoresist material, of desired configuration, to improve tracking without compromising HF detection. Likewise, by independently controlling the configurations of the respective exposures by the primary and secondary beams, the invention provides a mechanism to improve Figure of Merit in PR applications, as it does in dye-polymer applications, by optimizing ultimate three-dimensional pit and land geometry for improved tracking with HF detection that may be improved as well.
If applied to mastering of pre-grooved discs, such as MO, PC and CD-R/DVD-R, the dual-beam concept of the invention can likewise improve tracking without compromising groove reflectivity, by providing a mechanism whereby the geometry of the grooves, in radial section, can more precisely be controlled and optimized. In particular, the berms that typically line the tracking grooves in these recording structures, if created by the dye-polymer process, can be substantially reduced in height, improving both tracking and data retrieval.
An alternative embodiment, applicable in at least each of the foregoing contexts, incorporates a dithered secondary beam whose intensity is greater, and whose focused spot size is narrower, than the secondary beam configuration previously described. In this embodiment, the secondary beam is ditheredxe2x80x94oscillated rapidly in a radial direction, in respect to the discxe2x80x94while trough formation occurs. A variation of the latter embodiment comprises use of beam splitting to produce a plurality of overlapping, side-by-side non-dithering beams whose sum is a beam of essentially uniform intensity along all or a portion of its width. A further variation comprises beam dithering without beam splitting, where the single beam executes a complex dithering motion that is radial in respect to the disc (i.e., transverse, in respect to the longitudinal dimension of the formed features), and thus may produce pits and a trough simultaneously.
Also disclosed are methods by which the improved optical recording structure may be achieved. These include methods inherent in the various apparatus just described, as well as polymer cross-linking methods to provide an active layer less prone to berm formation and/or to formation of residual berms with undesirable radial width.
The invention may similarly be applied to non-disc structures. For example, in the case of optical data recording on cylindrical structures, the circumferentially spiral track corresponds to the essentially planar spiral track on an optical data recording disc surface. Here, the principles are virtually identical, except that the primary and secondary write beams are directed radially toward the rotational axis of the spinning structure, and dithering (or beam splitting), if applied, occurs in a direction parallel to the spin axis. Similarly, in optical recording on essentially two-dimensional data cards, where the write beam scans substantially parallel data tracks while the medium is not in motion, this dual-beam principal (with or without dithering or beam-splitting)xe2x80x94or the complexly-dithered, single beam conceptxe2x80x94may be applied with suitable configuration and mechanical refinements. Likewise, in three-dimensional data wafers, which essentially comprise individually readable, sandwiched layers of perhaps microscopically-thin data cards, the invention may be applied as in the case of data card embodiments, except that the writing process is conducted on multiple layers of the medium, where focusing and selection of the particular material(s) comprising the medium may be accomplished in a reasonably conventional manner, as may be further refined as development of data wafers progresses in the future.
The invention, in all its embodiments, can be applied equally in embodiments wherein each data mark is a bump or ridge extending above original surface level of the structure. In one sense, what appears as a pit or groove, when viewed from the one direction is a bump or ridge when viewed from the opposite direction (as is the CD-ROM/DVD-ROM case, where disc masters are first surface written and second surface read). Even more clearly, a disc stamper is the complementary image of the original master, and it therefore displays ridges or bumps where the master originally displayed grooves or pits, respectively. Generally, however, as is known to those with ordinary skill in the art, the conversion from pit/groove writing to bump/ridge writing depends only on conventional selection of parameters and/or waveforms, and the present invention, in variations within the basic concept, can be equivalently applied to either.
One such alteration, within the scope of the present invention, would comprise essentially inverting and otherwise appropriately processing the waveform (or waveforms) controlling the write beam, so that instead of causing expulsion of pit areas, expulsion would occur in land areas and, selectively, in areas between adjacent data tracks, thus preserving data bumps, relative to the now lowered land and inter-track areas. Other methods of raised feature creation will doubtless occur to those with ordinary skill in the art, based on the teachings herein, the published art and well-known principles of optical and circuit design.
The invention has been shown to be capable of virtually eliminating the berms typically surrounding data pits or (CD-R/DVD-R) tracking grooves in dye-polymer optical data recording. Indeed, this demonstrated reduction of berm height is sufficiently substantial that with careful parameter/material selection and experimentation within the scope of the disclosed invention and by application of reasonable skill in the art, it appears possible to actually eliminate these berms altogether, thus even further improving HF detection. Here, each side of the radial section of the pit would be a continuous curve whose slope is monotonically negative from the surface level (or another defined level) of the structure to the base of the pit on one side of the pit and monotonically positive from the base of the pit back to the surface level (or other defined level) on the opposite side, and the pit section would not display any substantial slope discontinuity from one side to the other side. The latter would mean that the ratio of berm height to pit depth would be zero. In the case of a bump embodiment of the latter, the ratio of bump height to side channel depth would be infinite. CD-R/DVD-R groove sectional shape would, of course, be quite similar, as would tracking ridge sectional shapes, with similar ratios.
All embodiments of the invention share a single goal: to improve Figure of Merit in optical recording structures. The invention accomplishes this goal by providing three-dimensional features of improved transverse sectional shape in the data layers of these structures. Thus, the invention, in all of its many embodiments, constitutes a unitary concept.
Certain of these embodiments of the invention will now be described in reference to the Drawing.